JP4808622B2 - Strain channel CMOS transistor structure having lattice-mismatched epitaxial extension region and source and drain regions and method of manufacturing the same - Google Patents

Description

  The present invention relates to the manufacture of semiconductor integrated circuits, and more particularly to strained channel complementary metal oxide semiconductors (CMOS) having lattice-mismatched epitaxial extension regions and source and drain regions. The present invention relates to a transistor structure and a manufacturing method thereof.

  Both theoretical and experimental studies have shown that applying sufficient stress to the conduction channel to cause distortion in the transistor's conduction channel can significantly increase the carrier mobility of the transistor. Yes. Stress is defined as the force per unit area. Strain is the change in dimension of an item relative to the original dimension when a force is applied in the direction of a particular dimension, for example, when a force is applied in the length direction of an item. , A dimensionless quantity defined as the change in length relative to the initial length. The strain can be a tensile strain or a compressive strain. In a p-type field effect transistor, when compressive longitudinal stress, that is, compressive stress in the length direction of the conduction channel, is applied, distortion of the conduction channel, which is known to increase the drive current of the PFET, occurs. However, when the same stress is applied to the NFET conduction channel, the NFET drive current decreases.

  It has been proposed to improve the performance of NFETs and PFETs by applying tensile longitudinal stress to the NFET conduction channel and applying compressive longitudinal stress to the PFET conduction channel. Such a proposal focuses on a masked process that masks the PFET portion of the chip and changes the material used for the shallow trench isolation region near the PFET conduction channel to apply the desired stress to the PFET conduction channel. It is matched. A separate step is then performed to mask the NFET portion of the chip and change the material used for the shallow trench isolation region near the NFET conduction channel to apply the desired stress to the NFET conduction channel. Other proposals include a masked process focused on adjusting the inherent stress present in the spacer features.

  Silicon germanium is a lattice mismatched semiconductor that is desirable for use in forming strained silicon transistor channels. Strain is created when a second semiconductor is grown on a single crystal of the first semiconductor when the first semiconductor and the second semiconductor are lattice mismatched to each other. Silicon and silicon-germanium are lattice mismatched to each other so that when one is grown on top of the other, tensile or compressive strain is created within each.

  Silicon-germanium having a crystal structure that matches the crystal structure of silicon is epitaxially grown on the silicon. However, since silicon germanium usually has a larger crystal structure than silicon, the epitaxially grown silicon germanium is compressed internally.

  In other proposals using strained silicon, the substrate includes a very thick silicon-germanium layer. Alternatively, the bulk substrate is made of single crystal silicon / germanium. Both the silicon germanium layer and the substrate are known as relaxed layers because the strain is released by the dislocations that form in the silicon germanium layer. When a single crystal silicon layer is epitaxially grown on the stress relaxation layer of single crystal SiGe, tensile strain is generated in the epitaxially grown silicon crystal layer. As a result, the electron mobility is improved, thereby improving the performance of the NFET.

  However, such a technique requires that the SiGe be stress relaxed, which requires that the SiGe layer be very thick, i.e., that the thickness be at least 0.5 to 1.0 [mu] m. It is difficult to improve the hole mobility. This is because in order to do so, it is necessary to increase the proportion of germanium in the SiGe layer. When the proportion of germanium is increased, excessive dislocations may occur in the SiGe crystal, which causes a yield problem. Furthermore, processing costs can be prohibitive.

  Other techniques such as graded Ge concentration methods, chemical mechanical polishing methods, etc. are also used to improve the quality of these films. However, high cost and high defect density are a problem with these techniques.

  Therefore, it is desirable to create strain in the channel region of the PFET without using a thick SiGe crystal region. It is desirable to use SiGe films epitaxially grown in the source and drain regions of the PFET to create the desired strain in the device channel region.

  Furthermore, it is desirable to form the SiGe film sufficiently thin in order to allow the SiGe film to apply a desired large stress and to prevent the SiGe film from becoming a stress relaxation film.

  Furthermore, it is desirable to create a compressive strain that increases hole mobility in the channel region of the PFET by growing SiGe epitaxial layers in the source and drain regions of the PFET.

  In addition, it is desirable to provide a process that forms higher source and drain regions that extend above the level of the gate dielectric and include lattice mismatched semiconductors to create the desired strain in the channel region of the PFET.

  Furthermore, it is desirable to provide a process that produces the desired strain in the channel region of the PFET but does not cause the same strain in the channel region of the NFET.

  Furthermore, a lattice mismatched semiconductor layer is formed in the source and drain regions of the PFET near the channel region of the PFET, and at the same time, the lattice mismatched semiconductor layer is not formed near the channel region of the NFET of the same integrated circuit. It is desirable to provide a structure and a method for manufacturing the same.

  Furthermore, a lattice mismatched semiconductor layer is formed in the extended region of the PFET near the channel region of the PFET, and at the same time, the lattice mismatched semiconductor layer is not formed in the region near the channel region of the NFET of the same integrated circuit. It would be desirable to provide such a structure and method for making the same.

  According to one aspect of the invention, an integrated circuit is provided having complementary metal oxide semiconductor (CMOS) transistors including a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET). Each of the NFET and the PFET has a channel region disposed in the first semiconductor single crystal layer, and the PFET channel region is applied with a first magnitude of stress. Has not been added. This stress is applied by a second semiconductor layer that is lattice mismatched with the first semiconductor. This second semiconductor layer is formed in the PFET source and drain regions at a first distance from the PFET channel region. The second semiconductor layer is also formed at a second distance from the NFET channel region in the NFET source and drain regions, where the second distance is greater than the first distance.

  In accordance with another aspect of the present invention, a method of manufacturing an integrated circuit including a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET) is provided. Each of the NFET and the PFET has a channel region disposed in the single crystal region of the first semiconductor, and a stress of the first magnitude is applied to the channel region of the PFET, but this is applied to the channel region of the NFET. The first magnitude of stress is not applied.

  According to a preferred aspect of the method, a gate conductor overlies the gate dielectric formed on the main surface of the first semiconductor single crystal region, over the first semiconductor single crystal region; A PFET gate stack and an NFET gate stack are formed, each having a first spacer including a first material formed on a sidewall of the gate conductor. A second spacer comprising a second material is formed on the sidewalls of the first spacers of the PFET gate stack and the NFET gate stack. Then, a portion of the second material from the second spacer of the PFET gate stack is relative to the first material while preventing the second material from being removed from the second spacer of the NFET gate stack. Selectively removed. Thereafter, a second semiconductor layer lattice-mismatched with the first semiconductor is grown on the exposed region of the single crystal region of the first semiconductor, and the PFET channel region has a first size. Although stress is applied, this first magnitude of stress is not applied to the channel region of the NFET. Source and drain regions are fabricated to complete the PFET and NFET.

  FIG. 1 shows a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET) according to one embodiment of the present invention. As shown in FIG. 1, the NFET 10 and the PFET 20 are manufactured in the single crystal semiconductor region 14 of the substrate 16. The substrate 16 may be a bulk substrate, or preferably a semiconductor-on-insulator substrate in which a relatively thin semiconductor single crystal region is formed on the insulating layer 18, such as a silicon-on-insulator (SOI) substrate. You can also. When a field effect transistor (FET) is formed on such an SOI substrate, the junction capacitance between the channel region of the transistor and the bulk substrate is eliminated, so that a switching operation that is faster than the case where the SOI substrate is not used is often achieved. The The substrate preferably includes the single crystal silicon region 14, and more preferably an SOI substrate having the single crystal silicon region 14 on the insulating layer 18.

See the fabrication of NFET and PFET transistors having channel regions disposed within a single crystal region of a substrate that is preferably substantially composed of a first semiconductor, such as silicon, as described in this and subsequent embodiments. To do. Since the first semiconductor is preferably silicon, the lattice-mismatched second semiconductor is preferably a different semiconductor such as silicon germanium or silicon carbide, and is silicon germanium (Si _x Ge _y ). It is more preferable. Where x and y are percentages and x + y is 100 percent. The range of variation between x and y can be quite large, illustratively y varies from 1% to 99%, in which case x results in a variation between 99% and 1%. .

However, the present invention is not limited to manufacturing transistors in pure silicon crystals. The single crystal region of the substrate 14 can consist essentially of a proportion of silicon germanium based on the first formula Si _x1 Ge _y1 , where x1 and y1 are percentages, x1 + y1 = 100%, The semiconductor layer consists essentially of different proportions of silicon germanium based on the second formula Si _x2 Ge _y2 , where x2 and y2 are percentages, x2 + y2 = 100%, x1 and x2 are not equal, y1 and y2 is not equal. A second semiconductor that is lattice mismatched with the first semiconductor is formed by epitaxial growth in the source and drain regions of the PFET near the channel region of the PFET, and at the same time, the second semiconductor that is lattice mismatched is NFET It is not formed near the channel region.

  FIG. 2 provides an aid in understanding the principles upon which the structures and methods of embodiments of the present invention are based. FIG. 2 is a graph showing the magnitude of compressive stress induced in a region of interest by a thin silicon germanium epitaxial layer laterally displaced from the single crystal silicon region of interest. The curve in FIG. 2 represents the magnitude of the compressive stress mapped against the lateral displacement from the edge of the region of interest for each different concentration percentage of germanium in the epitaxial layer.

  As shown in FIG. 2, the SiGe layer having a Ge percentage of 37.5% applies a stress of 350 MPa to the region of single crystal silicon at a lateral displacement of 10 nm. However, as the lateral displacement from the SiGe layer increases, the magnitude of the stress decreases rapidly. Looking at the same Ge percentage of 37.5%, the stress drops to 150 MPa with a lateral displacement of 30 nm. A SiGe layer having a smaller percentage is also shown in the graph. A SiGe layer having a Ge percentage of 6.25% applies a stress of 75 MPa to the region of single crystal silicon at a lateral displacement of 10 nm. However, the stress applied at a lateral displacement of 30 nm decreases to about 30 MPa. The remaining curves in the graph show that the stress induced in the channel increases with increasing Ge content.

  The embodiments described herein take advantage of the rapid decrease in stress associated with increased lateral displacement to form PFETs with strain-induced lattice mismatch source and drain regions near the channel region. On the other hand, an NFET is formed that does not have strain-induced lattice mismatch source and drain regions near the channel region.

The teachings of the present invention, it is to be understood to be applicable to the preparation of the composition _{Al A In B Ga C As D} P E other types of semiconductor transistors, such as III-V compound semiconductor having _{N F.} Here, A, B, C, D, E, and F represent percentages of the respective elements Al, In, Ga, As, P, and N in the semiconductor crystal, and these percentages are 100 in total. Gallium arsenic (GaAs), indium phosphide (InP), gallium nitrogen (GaN) and InGaAsP are common examples of such semiconductors.

  As shown in FIG. 1, the PFET 20 includes a channel region 22 disposed under a gate conductor, which preferably includes an underlayer of heavily doped polysilicon 26 in contact with a gate dielectric 27. The gate dielectric 27 is preferably a silicon dioxide layer thermally grown on the single crystal semiconductor region 14. A halo region 23 and an extension region 25 are preferably disposed adjacent to the source and drain regions 24 and near the channel region 22.

The polysilicon underlayer 26 of the gate conductor is preferably heavily doped to a concentration of about 10 ¹⁹ cm ⁻³ . In order to match the work function of the p-type conduction channel that is present when the PFET is turned on during operation, the polysilicon layer 26 of the PFET 20 preferably includes a p-type dopant such as boron. The gate conductor preferably further includes a low resistance portion 28 disposed over the polysilicon portion 26. The low resistance portion 28 has a much lower resistance than the polysilicon portion 26 and preferably comprises a metal or metal silicide, or both. In a preferred embodiment, the low resistance portion 28 includes a silicide formed by a self-aligned process (“salicide”), which is a suitable metal silicide including tungsten, titanium, and cobalt. However, suitable metals are not limited to these. More preferably, the silicide is a cobalt compound (CoSi ₂ ).

  The gate conductor may alternatively include a metal layer instead of the polysilicon layer in contact with the gate dielectric 27, which metal layer is replaced by a replacement gate after high temperature processing of the source and drain regions of the transistor is complete. It is preferable that it is formed as.

Single crystal silicon region 16 includes NFET 10 and PFET 20 separated from channel regions 122 and 22 of NFET 10 and PFET 20 by a first spacer 30 pair, a second spacer 32 pair, and a third spacer 34 pair, respectively. Source and drain regions 24 are formed. On the source and drain regions 24 of the NFET 10, a pair of higher source-drain regions 36 including a silicon germanium epitaxial layer 39 and a low resistance layer 40 are disposed. Above the source and drain regions 24 of the PFET 20, a pair of higher source-drain regions 36 including a silicon germanium layer 38 and a low resistance layer 40 are disposed. This low resistance layer may be a silicide, or “salicide”, formed in a self-aligned manner from a metal deposited on the silicon-germanium layers 38, 39 and later reacted with the silicon-germanium to form a silicide. preferable. The silicide can be a compound of a suitable metal including tungsten, titanium and cobalt. But not limited to those suitable metal. More preferably, the silicide is a cobalt silicide, that is, CoSi ₂ .

  As shown in FIG. 1, the silicon-germanium layer 38 extends laterally under the second and third spacers 32 and 34 of the PFET 20 to the sidewalls of the first spacer 30. In this way, the silicon-germanium epitaxial layer 38 is placed near the channel region 22 of the PFET and applies a compressive stress that can favor the hole mobility in the channel region 22. In order for the epitaxial layer 38 to apply a desired amount of stress to the channel region 22, the width of the first spacer 30 is preferably 10 nm or less.

  In contrast to PFET 20, epitaxial layer 39 of NFET 10 is laterally displaced from channel region 122 by a distance that spans at least the width of first and second spacers 30, 32. In this way, the silicon-germanium epitaxial layer 39 is not placed as close as possible to the NFET channel region 122 to adversely affect NFET performance.

  3-12 illustrate the steps of a CMOS manufacturing process according to one embodiment of the present invention. As a result of the processing based on this embodiment, a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET) are formed. In a PFET, a first magnitude stress is applied to the channel region by the lattice mismatched semiconductor layer. On the other hand, since the lattice-mismatched semiconductor layer is not located near the channel region of the NFET, the first magnitude stress is not applied to the channel region of the NFET. In this way, an increase in the carrier mobility of the PFET is achieved while maintaining the desirable performance of the NFET.

  FIG. 3 illustrates one stage of a process for forming PFETs and NFETs according to one embodiment of the present invention. As shown in FIG. 3, a PFET gate stack 44 and an NFET gate stack 45 are formed over the single crystal semiconductor region 14 of the substrate. The single crystal region 14 is substantially made of the first semiconductor material described above. PFET gate stack 44 consists essentially of silicon nitride, with gate dielectric 27 overlying single crystal region 14, gate conductor layer 26, preferably in contact with polysilicon, in contact with the gate dielectric. And a preferable insulating cap 50. NFET gate stack 45 consists essentially of silicon nitride, gate dielectric 27 overlying single crystal region 14, gate conductor layer 26, preferably in contact with gate dielectric 27, preferably comprising polysilicon. And a preferable insulating cap 50.

  In one embodiment, already at this stage, the gate conductors 26 of the PFET gate stack and NFET gate stack have the desired dopant type and concentration to provide the desired work function. For example, the PFET gate stack 44 has a p + doped gate conductor layer 26 and the NFET gate stack 45 has an n + doped gate conductor layer 26.

  Next, as shown in FIG. 4, NFET gate stack 45 is used as a mask to prevent implantation from penetrating too deeply into channel region 122 under NFET gate stack 45. Preferably, an extension implantation and a halo implantation into the active region of the single crystal region 14 adjacent to the stack 45 are performed. During such implantation, the active region adjacent to the PFET gate stack 44 is prevented by implantation, such as by a block mask 42, which preferably includes a photoresist material.

Next, as shown in FIG. 5, the block mask 42 is removed and a first pair of spacers 30 are formed on the sidewalls of the PFET gate stack 44 and the NFET gate stack 45. The spacer 30 is formed from a deposited nitride such as silicon nitride and is preferably thin, for example, a thickness of 3 nm to 20 nm, more preferably 5 nm to 15 nm, and most preferably about 10 nm.

  Next, as shown in FIG. 6, the PFET gate stack 44 is used as a mask to prevent implantation from penetrating too deeply into the channel region 22 below the PFET gate stack 44. It is preferred that extension implantation and halo implantation into the active region of the single crystal region 14 adjacent to the stack 44 is performed. During such implantation, the active region adjacent to the NFET gate stack 45 is prevented by implantation, such as by a block mask 43, which preferably includes a photoresist material.

  Thereafter, the block mask 43 is removed and a thick conformal material layer 46 is deposited over the PFET gate stack 44 and the NFET gate stack 45, as shown in FIG. The conformal material layer 46 should effectively be an insulating layer rather than a conductive layer or a semiconductive layer. The conformal material layer 46 preferably comprises an oxide, preferably silicon dioxide, and is preferably deposited at a low temperature from a precursor such as tetraethylorthosilicate (TEOS). Hereinafter, the material of the layer 46 is referred to as “oxide”.

  Next, as shown in FIG. 8, an additional spacer or third spacer 48 pair, preferably comprising a nitride material, and more preferably silicon nitride, is connected to the PFET gate stack 44 and the NFET gate stack 45. Formed on both oxide layers 46. This process deposits a conformal layer of silicon nitride and then etches the structure vertically, such as by reactive ion etching (RIE), leaving spacers 48 on the sidewalls of oxide layer 46, and conformal nitridation from the horizontal plane. It is preferably carried out by allowing the layer to be removed.

  Next, as shown in FIG. 9, after the nitride spacers 48 are in place, the oxide layer 46 on the top surface of the structure is formed into a PFET gate stack 44, such as by RIE selective to nitride. And to the level of the insulating cap 50 of both the NFET gate stack 45. During such etching, the oxide layer 46 is also removed from the regions of the single crystal region 14 that extend beyond the nitride spacers 48 of the respective PFET gate stack 44 and NFET gate stack 45. During such etching, the nitride spacer 48 protects the sidewalls of the structure from being etched, and the insulating cap 50 damages or etches the gate conductor 26 of the PFET gate stack and NFET gate stack, or otherwise. Protect from both.

  Thereafter, as shown in FIG. 10, a block mask 52 is again applied over the area containing the NFET gate stack 45, leaving the PFET gate stack 44 exposed. Block mask 52 preferably comprises a photoresist material. After the block mask 52 is in place, the oxide layer 46 deposited on the PFET gate stack 44 is undercut, such as by an isotropic wet chemical etch selective to nitride. As a result, the oxide layer 46 having the appearance shown in FIG. 10 is obtained. As a result of this etching, main surface 54 of single crystal semiconductor region 14 is exposed.

  Thereafter, as shown in FIG. 11, a single crystal layer of a second semiconductor that is lattice-mismatched with the first semiconductor is epitaxially grown on the main surface of the single crystal semiconductor region 14. As described above with respect to FIG. 1, regardless of whether the single crystal semiconductor region 14 includes germanium, the second semiconductor is silicon germanium having a germanium percentage that is higher than the germanium percentage of the single crystal semiconductor region 14. Preferably there is. In the PFET region, this layer 38 is formed under the undercut portion 56 of the oxide layer 46 so that the layer 38 laterally separated from the channel region 22 only by the first nitride spacer 30 is A compressive stress is applied near the channel region 22 of the PFET 20.

  On the other hand, in the NFET, since the compressive stress hinders the electron mobility of the NFET, the silicon-germanium layer 39 is not formed near the gate conductor 26. Therefore, the compressive stress that the layer 39 applies to the channel region 122 of the NFET. Is not as great as the compressive stress applied to the channel region 22 of the PFET. However, if the lattice-mismatched semiconductor layer that induces this stress is displaced a sufficient distance from the channel region 122 of the NFET 10, its compressive stress can be tolerated as described above with respect to FIG. . In addition, the spacer 30 and oxide layer 46 parameters can be adjusted to apply a small counter stress that improves electron mobility in the NFET. Such reverse stress will be applied as a small tensile stress to counter the effect of the small compressive stress applied by the silicon-germanium layer 39 into the NFET channel region 122.

  The final processing stage of this embodiment is shown in FIG. During this processing step, the PFET gate stack structure 44 including the gate conductor 26, the first spacer 30, the second spacer 32, and the third spacer 34 is used as a mask in the single crystal region 14. Source and drain regions 24 of PFET 20 are implanted. Meanwhile, the region of NFET 10 is protected from this implantation by a block mask (not shown). Preferably in a separate implantation step, into the single crystal region 14 using an NFET gate stack 45 comprising a gate conductor 26, a first spacer 30, a second spacer 32 and a third spacer 34 as a mask. The source and drain regions 24 of NFET 10 are implanted. Meanwhile, PFET 20 is protected from this implantation by a block mask (not shown). Thereafter, the implanted source and drain regions 24 may be annealed and a high temperature process may be performed to drive the implanted dopant to the desired depth and lateral dimensions.

  At this time, nitride insulation cap 50 is removed from PFET gate stack 44 and NFET gate stack 45. A silicide-forming metal is then deposited over the structure shown, which is then subjected to high temperature processing with the semiconductor material of the polysilicon gate conductor 26 in contact therewith and the silicon germanium layers 38 and 39 also in contact therewith. It is preferred to react to form self-aligned silicide (“salicide”) 40. Alternatively, following the high temperature anneal of the source and drain regions 24, the nitride insulation cap 50 and the polysilicon gate conductor 26 between the spacers 30, 32 are removed, such as by RIE selective to nitride and oxide. A metal replacement gate can also be formed at that position. In such an alternative process, the previously formed gate dielectric preferably functions as an etch stop layer or sacrificial layer of polysilicon RIE. After RIE removal of polysilicon gate 26, the initially formed gate dielectric is removed due to damage of that layer during RIE. Thereafter, a second gate dielectric 27 is deposited at the location previously occupied by the removed first gate dielectric. A metal gate conductor is then deposited in the opening thereby formed between the spacers 30, 32 as a conformal layer of single crystal silicon-germanium layers 38, 39. In such a method, the metal replacement gate is formed after the processing of PFET 20 and NFET 10 is substantially completed.

  Another embodiment of a PFET 220 and an NFET 210 formed in accordance with the present invention is shown in FIG. In this embodiment, the silicided higher source and drain regions 224 are displaced from the channel region of NFET 210 by the desired distance, and the silicided higher source and drain regions 224 are displaced from the channel region of PFET 220 by the desired distance. Therefore, four pairs of spacers are used in the NFET 210. As shown in FIG. 12, the lattice-mismatched semiconductor layer 238 of the PFET is formed as a higher layer close to the channel region 322 of the PFET that is in contact with the single crystal semiconductor region 214. In NFET 210, the lattice-mismatched semiconductor layer 239 is formed as a higher layer, but this layer is near the channel region 222 of the NFET 210 due to the presence of an additional spacer 231 between the layer 239 and the channel region 222. Not formed. In this embodiment, layers 239 and 238 apply different amounts of compressive stress to the NFET and PFET channel regions 222 and 322.

  In this embodiment, the amount of strain created in the channel region of NFET 210 can be adjusted based on the width 240 of the second spacer 231. As discussed above, the greater the lateral displacement of the layer 239 from the channel region 222 of the NFET 210, the less strain is created in the channel region 222 of the NFET. A small strain has a smaller negative effect on the electron mobility of the NFET 210 than a large strain. In this embodiment, such a small strain can be achieved by using an appropriately sized spacer made substantially of a material such as silicon nitride.

  The spacer 231 has a width 240 determined by the thickness of the deposited conformal silicon nitride material. If even smaller strain is required in the channel region 222 of the NFET 210, the silicon nitride spacer 231 can be made thicker by depositing this silicon nitride layer thicker.

  The manufacturing steps of this embodiment will now be described with reference to FIGS. As shown in FIG. 13, the PFET gate stack 244 and the NFET gate stack 245 are each preferably a gate-grown oxide on a single crystal semiconductor region 214 of a substrate, such as a single crystal silicon region. A polysilicon gate 226 overlying 227 is included. Above the polysilicon gate 226 is an insulating cap 250. A pair of first spacers 230 are formed on the sidewalls of the polysilicon gate 226 after patterning and etching the gate stack structures 244, 245. These first spacers 230 are preferably thin and preferably have a thickness of 3 nm to 20 nm, more preferably 5 nm to 15 nm, and most preferably about 10 nm.

  After forming the spacers 230, halo and extended ion implantations into the source and drain regions (not shown) of the PFET 220 and NFET 210 adjacent to these spacers preferably place a block mask over the NFET region, This is done by implanting into the PFET region and then applying a block mask to the PFET region while implanting into the NFET region. Thereafter, as shown in FIGS. 14 and 15, a pair of second spacers 231 is formed on the side walls of the pair of first spacers 230. This is performed by depositing a conformal material such as silicon nitride and then etching the structure vertically by RIE or the like to obtain the structure shown in FIG.

  Thereafter, a block mask 243 is applied over the NFET gate stack 245 and its adjacent regions, as shown in FIG. The second spacer 231 is then removed from the PFET gate stack 244. Next, as shown in FIG. 17, a layer of silicon germanium 238 is selectively grown on the single crystal region 214 of the substrate. Due to the presence of the second spacer 231 on the sidewalls of the NFET gate stack 245, the silicon germanium layer 238 is NFET more than the distance to the PFET channel region 322 (eg, the width 240 of the spacer 231). The channel region 222 is displaced laterally. In this way, increased hole mobility in the PFET is achieved without significantly affecting the electron mobility in the NFET.

  Next, as shown in FIG. 18, additional spacers 232 and 234 are formed. These spacers 232 and 234 are used to separate the final higher step silicided source and drain regions 224 (FIG. 12) from the NFET and PFET channel regions 222 and 322. These spacers 232 and 234 preferably include nitride and oxide, respectively. The spacer 232 is preferably made of nitride, and the spacer 234 is preferably made of oxide. During this process, an additional RIE etch is performed to obtain the structure shown in FIG. Finally, a self-aligned silicide layer 224 as shown in FIG. 12 is formed in the region of layer 238 that is not covered by gate stacks 244 and 245.

  Other embodiments of the present invention are shown in FIGS. In contrast to the embodiment shown with respect to FIGS. 12-18, in this embodiment, the silicon germanium layer 338 does not form part of the NFET structure 310, as shown in FIG. Layer 338 is disposed only on PFET structure 320. In this way, compressive stress is applied to the channel region 422 of the PFET 320 but not to the channel region 423 of the NFET 310.

The process for manufacturing PFET 320 and NFET 310 is shown in FIGS. As shown in FIG. 20, PFET gate stack 344 and the NFET gate stack 345 is formed, each of which overlap with reportage over it is preferred gate dielectric 327 formed of an oxide thermally grown Rishirikon - It includes a gate 326, an insulating cap 350 that preferably includes silicon nitride, and a pair of first spacers 330 that preferably include silicon nitride. At this time, a halo implant and an extended implant can be performed.

  Thereafter, a conformal material layer 360 is deposited and then patterned to cover only the active region of the single crystal semiconductor region 314 adjacent to the NFET gate stack 345. Such conformal material layers can illustratively be oxides or nitrides, or combinations thereof. The conformal material layer 360 preferably includes a nitride such as silicon nitride. Thereafter, a lattice mismatched semiconductor 338 such as silicon germanium is epitaxially grown on the exposed active region of the single crystal region 314 adjacent to the PFET gate stack 344.

  Thereafter, as shown in FIG. 21, an additional plurality of insulating layers are deposited on the conformal layer 360, and these layers are then etched vertically, such as by RIE, to form spacers 331 and additional spacers 332, 334. For example, it is formed by the method described above with reference to FIG. Thereafter, as shown in FIG. 19, self-aligned silicide regions 324 and 424 are preferably formed in both PFET 320 and NFET 310, for example, in the manner described above with respect to FIG.

  While the invention has been described in terms of preferred embodiments thereof, those skilled in the art will recognize that many changes and modifications may be made without departing from the true scope and spirit of the invention which is limited only by the appended claims. Let's understand.

FIG. 2 illustrates a PFET and an NFET according to one embodiment of the present invention. FIG. 6 is a graph showing the magnitude of compressive stress induced in a single crystal silicon region involved by a thin silicon germanium epitaxial layer. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 4 illustrates a stage in the manufacture of PFETs and NFETs according to one embodiment of the invention. FIG. 5 shows a stage in the manufacture of PFETs and NFETs according to another embodiment of the invention. FIG. 5 shows a stage in the manufacture of PFETs and NFETs according to another embodiment of the invention. FIG. 5 shows a stage in the manufacture of PFETs and NFETs according to another embodiment of the invention. FIG. 5 shows a stage in the manufacture of PFETs and NFETs according to another embodiment of the invention. FIG. 5 shows a stage in the manufacture of PFETs and NFETs according to another embodiment of the invention. FIG. 5 shows a stage in the manufacture of PFETs and NFETs according to another embodiment of the invention. FIG. 6 shows a stage in the manufacture of PFETs and NFETs according to yet another embodiment of the invention. FIG. 6 shows a stage in the manufacture of PFETs and NFETs according to yet another embodiment of the invention. FIG. 6 shows a stage in the manufacture of PFETs and NFETs according to yet another embodiment of the invention.

Claims (18)

  An integrated circuit structure having complementary metal oxide semiconductor (CMOS) transistors including a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET), wherein each of the NFET and the PFET is a first A second semiconductor layer having a channel region disposed in a single crystal layer of the semiconductor and having a lattice mismatch with the first semiconductor causes a stress of a first magnitude in the channel region of the PFET. Although not added to the channel region of the NFET, the layer of the second semiconductor is at a first distance from the channel region of the PFET in the source and drain regions of the PFET. Where the layer of the second semiconductor is in the source and drain regions of the NFET and the channel of the NFET. Le region is also formed at a second distance, said second distance is greater than said first distance integrated circuit structure. The first semiconductor and the second semiconductor are silicon-containing semiconductor materials having a composition based on the formula Si _x Ge _y , x and y are percentages, and the composition of the first semiconductor is x = 100, y = 0 to x = 1, y = 99, and the composition of the second semiconductor is x = 99, y = 1 to x = 1, y = 99, The integrated circuit structure of claim 1, wherein x of the semiconductor is always smaller than x of the first semiconductor.   The single-crystal region of the first semiconductor has a major surface defined by a level of gate dielectric formed on the channel region of the NFET and the PFET, and the layer of the second semiconductor is The integrated circuit structure according to claim 1, wherein the integrated circuit structure is formed on the main surface. The first semiconductor is made of a semiconductor selected from the group consisting of silicon, silicon germanium and silicon carbide, and the second semiconductor is selected from the group consisting of silicon, silicon germanium and silicon carbide, The integrated circuit structure according to claim 1, wherein the integrated circuit structure is made of a semiconductor different from the first semiconductor. Made from the first semiconductor starvation silicon, consisting of the second semiconductor starve silicon-germanium, the integrated circuit structure of claim 1. Said first semiconductor, a silicon-germanium based on the first formula _Si x1 _{Ge y1,} a x1 and y1 are percentages, x1 + y1 = a 100%, y1 is at least 1 percent, the second semiconductor, a silicon-germanium based on the second formula _Si x2 _{Ge y2,} x2 and y2 are percentages, x2 + y2 = a 100%, y2 is at least 1 percent, unequal x1 and x2, y1 The integrated circuit structure of claim 1 wherein y2 is not equal to y2.   The integrated circuit structure of claim 1, wherein the stress is a compressive stress. An integrated circuit having complementary metal oxide semiconductor (CMOS) transistors including a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET) each having a channel region disposed within a single crystal silicon region of the substrate A structure, wherein the PFET source and drain regions are disposed at a first distance from the PFET channel region, the NFET source and drain regions, and the NFET channel region from the NFET . by a raised lattice-mismatched semiconductor layer made of divorced germanium disposed at a larger second distance greater than the first distance, said channel region of said PFET but is added first stress, It is not added to the channel region of the NFET and the silicon gate Maniumu has a composition based on the formula _Si x Ge _y, x and y is a percentage that is at least 1%, respectively, the integrated circuit structure x + y is 100 percent. A method of manufacturing an integrated circuit structure including a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET) comprising:
A gate conductor overlying a gate dielectric formed on a main surface of the first single crystal semiconductor region on the first single crystal semiconductor region having the first composition; and on a side wall of the gate conductor. Forming a PFET gate stack and an NFET gate stack each having a first spacer comprising a formed first material;
Forming a second spacer comprising a second material on sidewalls of the first spacer of the PFET gate stack and the NFET gate stack;
The portion of the second material from the second spacer of the PFET gate stack is removed from the second spacer of the PFET gate stack while preventing the second material from being removed from the second spacer of the NFET gate stack. Selective removal of materials,
Thereafter, an epitaxial single crystal semiconductor layer having a second composition and a lattice mismatch with the first single crystal semiconductor region is grown on the exposed area of the first single crystal semiconductor region. ,
Forming source and drain regions of the PFET including at least a portion of the epitaxial single crystal semiconductor layer;
Forming source and drain regions of the NFET including at least a portion of the epitaxial single crystal semiconductor layer ;
Including methods. 10. The method of claim 9 , comprising forming the PFET source and drain regions in the layer of the epitaxial single crystal semiconductor above the level of the major surface. 10. The method of claim 9 , comprising forming the NFET source and drain regions in the layer of the epitaxial single crystal semiconductor above the level of the major surface. Self-aligned silicide (salicide) to at least one selected from the source and drain regions of the PFET, the source and drain regions of the NFET, the gate conductor of the PFET gate stack, and the gate conductor of the NFET gate stack 10. The method of claim 9 , further comprising: The method of claim 9 , wherein the first composition is silicon, the second composition is silicon germanium, and the silicon germanium has a germanium content of at least 1 percent. Prior to forming the first spacer, the PFET and NFET are ion implanted into the region of the first single crystal semiconductor region masked by the gate conductor of the PFET gate stack and the NFET gate stack. The method of claim 9 , further comprising forming an implantation region that is self-aligned to the channel region. Prior to removing a portion of the second material from the second spacer, further comprising forming a third spacer on a sidewall of the second spacer, wherein the third spacer comprises the PFET The method of claim 9 , wherein a spacing between source and drain regions and the channel region of the PFET is defined. A mask covers and protects the first and second spacers formed on the NFET gate stack when removing the second material portion of the second spacer from the PFET gate stack. 10. The method of claim 9 , further comprising: A method of manufacturing an integrated circuit structure including a p-type field effect transistor (PFET) and an n-type field effect transistor (NFET) having a channel region disposed in a single crystal region made of a first semiconductor,
A first portion of a second semiconductor that is lattice-mismatched with the first semiconductor and from the channel region of the PFET such that a first magnitude of stress is applied to the channel region of the PFET. Forming source and drain regions of the PFET having the first portion disposed at a first distance;
In the second portion made of the second semiconductor, a stress having a second magnitude smaller than the first magnitude is applied to the channel area of the NFET from the channel area of the NFET. Forming the source and drain regions of the NFET having the second portion disposed at a second distance greater than a distance of one;
Including methods. The single crystal region has a major surface defined by a level of gate dielectric formed on the channel regions of the PFET and the NFET, and the first and second portions are on the major surface. The method of claim 17 , comprising forming.

Images